WO2009046713A1

WO2009046713A1 - Method and apparatus for controlling a dual fuel compression ignition engine

Info

Publication number: WO2009046713A1
Application number: PCT/DK2007/050145
Authority: WO
Inventors: Stefan Mayer
Original assignee: Man Diesel, Filial Af Man Diesel Se, Tyskland
Priority date: 2007-10-08
Filing date: 2007-10-08
Publication date: 2009-04-16
Also published as: CN101589213A; KR20100074084A; CN101589213B; KR101363942B1; CH698352B1; JP2010514967A; JP5053385B2

Abstract

A compression ignition internal combustion engine has combustion sensors for detecting combustion of fuel in the cylinders, and at least one electronic control unit receiving signals from the combustion sensors. A fuel mixing unit is connected with at least a first source of a first fuel and a second source of a fuel conditioner. Based on a detected start of ignition of fuel in the individual cylinder compared with start of fuel injection in said cylinder, the at least one electronic control unit controls the mixing ratio of the first fuel and the fuel conditioner in the fuel mixture delivered by the mixing unit to the fuel injectors, so that the ignitability of the fuel mixture is decreased when the ignition delay is at a predetermined value for ignition delay.

Description

METHOD AND APPARATUS FOR CONTROLLING A DUAL FUEL COMPRESSION IGNITION ENGINE

5 The present invention relates to a compression ignition internal combustion engine, such as a two-stroke crosshead Diesel engine or a four-stroke Diesel engine, comprising cylinders provided with fuel injectors for injecting fuel directly into combustion chambers in the cylinders, combustion sensors for detecting combustion of fuel in the cylinders,

10 and at least one electronic control unit receiving signals from the combustion sensors.

An engine of this kind is known from the market as a ME engine of MAN Diesel where the combustion sensors are e.g. exhaust gas temperature sensors. The engine is typically operated on a single fuel, such

15 as heavy fuel oil, or on a main fuel and a pilot fuel, e.g. gas as main fuel and oil as pilot fuel causing the initiation of the compression ignition in the combustion chamber. When two types of fuel are used simultaneously, such as in the case of main fuel and pilot fuel, the relative amounts spent of the two fuel types are more or less static, or change

20 very slowly over numerous combustion cycles.

For many years there have been developments in direction of obtaining lower specific fuel oil consumption, viz. less fuel consumed per kWh of produced power. And for the latest decade there have not been big achievements in this direction, mainly because large Diesel engines

25 are operating quite close to the theoretically obtainable efficiency.

According to the present invention it is a desire to optimize the fuel consumption in an economical sense so that costs of operating the engine are lowered.

With a view to this the engine according to the present inven-

30 tion is characterized in that a fuel mixing unit is connected with at least a first source of a first fuel and a second source of a fuel conditioner, that based on a detected start of ignition of fuel in the individual cylinder compared with start of fuel injection in said cylinder, the at least one electronic control unit controls the mixing ratio of the first fuel and the fuel conditioner in the fuel mixture delivered by the mixing unit to the fuel injectors, so that the ignitability of the fuel mixture is decreased when the ignition delay is at a predetermined value for ignition delay.

The electronic control unit can on basis of the detected start of ignition and on the start of the fuel injection determine a time difference representative of the actual ignition delay occurring at the combustion in the cylinder. The electronic control unit compares this time difference with a predetermined value for ignition delay, and if the time difference is nearly as short as the predetermined value for ignition delay then the control unit controls the mixing ratio so that the ignitability of the fuel mixture is decreased. The electronic control unit thus controls the fuel mixture in a manner increasing the use of fuel of low quality.

The detection of the start of ignition of fuel in the individual cylinder ensures a quick adjustment of the mixing ratio, because the elec- tronic control unit receives the information from the individual combustion process in the cylinder as the combustion is in progress. The electronic control unit may consequently adjust the mixing ratio before the next fuel injection is performed in the same cylinder. This makes the detection of the necessity to adjust the mixing ratio much faster than de- tection based on a temperature measurement in the exhaust gas where the measurement is actually averaged over many combustion cycles.

The fast detection of changes in the ignition delay makes it possible for the control unit to effectively compensate for ignition delay changes that are not related to the fuel composition, but instead caused by other factors such as the engine load, the humidity in the inlet air, the temperature of the cylinder parts (cylinder liner, cylinder cover and piston are colder when the engine has just been put into operation), the oxygen content in the inlet air, etc. Such changes cannot be avoided, and it has hitherto been considered a requirement to utilize fuel of a suf- ficiently high quality to ensure proper combustion at adverse engine operating conditions. The immediate detection of changes in the actual ignition delays makes it safe to utilize a fuel mixture of very low quality when the engine operating conditions are favourable, because the electronic control unit regulates the fuel mixture into a higher quality, should this be needed because of a change in the operating conditions.

The electronic control unit adds fuel of low quality to the fuel mixture when the detected ignition delay is as short as a predetermined value for ignition delay. The increase in fuel of low quality in the fuel mixture results in an increase in the ignition delay, and a lowering of the costs for the consumed fuel mixture.

In a preferred embodiment the first fuel is a fuel of lower costs than the fuel conditioner, which is a fuel of better quality than said first fuel. When the fuel constituents available have these qualities the elec- tronic control unit controls the fuel mixture to contain less of the higher- quality fuel conditioner, which is more expensive, and more of the first fuel, and this results in a cost saving of the fuel consumed. If the electronic control unit has adjusted the fuel mixture so that only fuel of the poorest quality is used, viz. the fuel mixture consists of only the first fuel, then the engine has to operate with the ignition delay resulting from this fuel, even if this value for the ignition delay should be smaller than the predetermined value for ignition delay. This will not present any problem to the engine, because it is capable of running on a fuel more readily ignitable than the poorest fuel, but the full potential for cost sav- ing may not be achieved. Information of such a situation can be utilized at the next purchase of fuel where one can obtain cheaper fuel of an even lower quality than the previously purchased first fuel.

In a preferred embodiment the first fuel has a CCAI-value that is higher than the CCAI-value of the fuel conditioner. The Calculated Car- bon Aromatic Index value (CCAI-value) of the fuel is a value dependent on the viscosity and the density of the fuel and can be taken as an indication of the ignition properties of the fuel oil, in the sense that a lower CCAI-value represents a fuel oil that is more readily ignitable than a fuel oil having a higher CCAI-value. In particular for internal combustion engines operating as main engines in ships the first fuel and the fuel conditioner may both be heavy fuel oils. These products are readily available for bunkering in ships, and the ability of the engine to run with heavy fuel oils of different qualities allows the ship operator to purchase and bunker very cheap - or the cheapest available - heavy fuel oil in conjunction with a heavy fuel oil of better quality.

In an embodiment the consumption of fuel conditioner is minimized by setting the predetermined value for ignition delay near a pre- determined maximum limit for ignition delay. In case the first fuel is of such a poor quality that the engine cannot run satisfactorily only on this fuel because the ignition delay would be unacceptably long, then it may be relevant to provide the electronic control unit with a predetermined maximum limit for ignition delay. And when such a value is provided it becomes possible to set the predetermined value for ignition delay near said maximum limit. The advantage of such a setting is that the electronic control unit will keep the mixture of the fuel close to the lowest fuel quality capable of maintaining the engine in operation, and consequently the fuel costs will be as low as possible for the fuel types avail- able for use in the engine.

In an embodiment only the first fuel is supplied to the fuel injectors when the engine is operating at 100% engine load. At 100% engine load the engine is more tolerant to fuel of low quality and this is utilized by supplying the engine with the lowest quality of fuel, and thus the cheapest fuel.

In a preferred embodiment the engine speed at 100% engine load is in the range from 45 rpm to 175 rpm. The ability to utilize poor quality fuel is more pronounced in engines running at low speed because there is firstly a longer period available for the individual combustion process because the engine cycle lasts longer when the engine speed is lower, and secondly the low speed engines are typically very large powered engines where it really matters to use a lower quality fuel because the fuel consumption is high. Some of the engines running at very low speed are engines for container ships, and such an engine may have a power of e.g. 75.000 kW or more.

As explained in the above the fuel conditioner can be a fuel that is more readily ignitable than the first fuel. However, it is within the scope of the present invention also possible that the first fuel is a heavy fuel oil and the fuel conditioner is water. Admixture of water to the fuel increases the ignition delay (decreases the ignitability).

The present invention also relates to a method of blending a fuel mixture that is injected by fuel injectors directly into combustion chambers in cylinders of a compression ignition internal combustion engine, such as a two-stroke crosshead Diesel engine or a four-stroke Diesel engine. Based on a detected start of ignition of fuel in the individual cylinder and compared with start of fuel injection in said cylinder, the mixing ratio of at least a first fuel and a fuel conditioner in the fuel mixture delivered to the fuel injectors is adjusted to maintain the ignition delay of the fuel mixture longer than a predetermined value for ignition delay. The advantages of this in view of saving of costs for fuel have been explained in the above description, and reference is made thereto.

Preferably, the mixing ratio is adjusted so to varying engine operating conditions that the costs of the fuel mixture are minimized. The method consequently involves a running adaptation of the fuel mixture delivered to the fuel injectors, to the current and actual operating conditions of the engine.

Examples of the present invention and embodiments thereof are in the following described in more detail with reference to the highly schematic drawing, in which

Fig. 1 is a general end view of an engine in accordance with the present invention,

Fig. 2 is an example of a fuel supply system for an engine according to the invention, Fig. 3 is an illustration of combustion sensors of optical type,

Fig. 4 is a more detailed illustration of one combustion sensor of optical type,

Fig. 5 depicts the end of an optical fibre of the combustion sensor in Fig. 4, and Fig. 6 is an illustration of the effects of shaping the end of the optical fibre with different angles.

A compression ignition internal combustion engine according to the present invention may be a two-stroke crosshead diesel engine as illustrates in Fig. 1. Such an engine can e.g. be of the make MAN Diesel and the type MC or ME, or of the make Wartsila of the type Sulzer RT- flex or Sulzer RTA, or of the make Mitsubishi Heavy Industries. An engine of this type is a large engine typically used as a main engine in a ship or as a stationary engine in a power plant. The cylinders can e.g. have a bore in the range from 25 cm to 120 cm, and the engine can e.g. have a power in the range from 3000 kW to 120.000 kW. The engine speed is typically in the range from 40 rpm to 250 rpm. The engine according to the invention can alternatively be a four-stroke diesel engine of an engine speed e.g. in the range from 300 rpm to 1400 rpm, and an engine power e.g. in the range from 1300 kW to 30.000 kW. The compression ignition internal combustion engines according to the present invention are typically capable of using heavy fuel oil as fuel.

The engine of Fig. 1 has a cylinder with a cylinder liner 1 mounted in a cylinder section 2 of an engine frame 3. An exhaust valve housing 4 is mounted in a cylinder cover 5 and an exhaust gas duct 6 extends from the individual cylinder to an exhaust gas receiver 7 common to several or all cylinders. In the exhaust gas receiver pressure variations caused by the exhaust gas pulses emitted from the exhaust gas ducts are equalized to a more even pressure, and one or more tur- bochargers 8 receive exhaust gas from the exhaust gas receiver 7 and deliver compressed air to a scavenge air system comprising at scavenge air receiver 9 which, like the exhaust gas receiver, is an elongated pressure vessel.

In Fig. 2 cylinder section 2 is illustrated with only a single cylin- der, but the engine has a plurality of cylinders, such as from 4 to 15 cylinders, when it is a two-stroke engine, and such as from 4 to 20 cylinders when it is a four-stroke engine. Each cylinder is associated with one or more fuel dosing devices 10. The fuel dosing device may be a fuel pump, and in that case the fuel system need only deliver fuel to the fuel dosing device at a relatively low feeding pressure in a fuel feeding pipe 11, such as a pressure in the range from 2 bar to 15 bar. Alternatively, the fuel dosing device may be a valve or a valve in connection with a metering device, and the fuel feeding pipe is then a high-pressure pipe in which the fuel is at a pressure higher than the injection pressure, such as a feeding pressure in the range of 500 bar to 1500 bar. Such a fuel system is called a common-rail system. In either case, the fuel dosing device 10 is connected to fuel feeding pipe 11 by a branch conduit 12 with a valve 13 that is maintained in open position during normal engine operation. Fuel dosing device 10 is connected to fuel injectors 14 via high-pressure fuel conduits 15. A return conduit 16 leads from the injectors to a fuel return line 17. The number of injectors per cylinder depends on the power of the cylinder. In smaller engines a single injector may be sufficient to inject the amount of fuel required for one combus- tion process, whereas in larger, more powerful engines two or three injectors may be required. When several injectors are provided per cylinder, there may be one fuel dosing device 10 per injector, or the several injectors may share one fuel dosing device as illustrated in Fig. 2.

In the individual cylinder a piston 18 is mounted on a piston rod 19 that is connected with a crank pin on a crankshaft via a crosshead and a connecting rod (not illustrated). The piston is connected directly to the connecting rod if the engine is a four-stroke engine. The fuel injector injects the fuel into a combustion chamber 20 where it auto-ignites because of the high temperature in the air above the piston. The high tem- perature is present because the piston has compressed the inlet air during the upward compression stroke. The ignition type is thus compression ignition, which is different from spark ignition utilized in engines working according to the Otto-cycle.

When the injector atomizes or injects the fuel into the com- pressed air in the combustion chamber then the fuel is heated by the hot air, and after some time - the ignition delay - the fuel ignites and begins to combust. The ignition delay begins with the injection of fuel. This starting point in time can be obtained either by actual measurement or based on the timing of the signal to activate fuel dosing device 10. A sensor mounted at the fuel injector 14 may perform the actual measurement and send a signal to an electronic control unit 21 via a wire 22. The sensor measuring the start of injection may be a magnetic sensor activated by the displacement of a needle in the injector, or a pressure sensor activated by the same displacement or by a high pres- sure in the fuel channel of the atomizer, or a vibration sensor connected to the back end surface of the atomizer. When the valve seat in the fuel injector opens and fuel flows forward and hits the end of the bore of the atomizer a pressure impulse is created that can be detected by the vi- bration sensor.

The starting point can alternatively be estimated with outset in the moment the fuel dosing device is activated. For a particular engine it is possible to determine the delay occurring between activation of the fuel dosing device and the actual start of injection of fuel into the com- bustion chamber. The electronic control unit 21 can then calculate the starting point as the point in time the fuel dosing device is activated plus the predetermined delay period. This determination of the starting point in time is very simple to implement, in particular when the electronic control unit also sends the activation signal to the fuel dosing device. There will naturally be a minor inaccuracy in this determination if the fuel properties vary significantly from the properties of the fuel used at the determination of the delay period. It is possible to compensate for this by allowing manual modification of the delay period, e.g. by entering a positive or negative value for a correction factor to be added to the predetermined delay period.

The ignition delay ends at the moment the fuel actually ignites in the combustion chamber. This ending point in time can be determined in various manners by using a combustion sensor detecting combustion of fuel in the cylinders. In one embodiment the combustion sensor is a pressure sensor detecting the rise in cylinder pressure caused by the combustion. Such a pressure sensor can be mounted in the sidewall of the cylinder liner at a position above the upper piston ring, when the piston is in the top dead centre position, or can be mounted in the cylinder cover. Alternatively, the pressure sensor can be a strain gauge device measuring the tension in a cover stud that clamps cylinder cover 5 to cylinder section 2 which tension can be converted to a pressure value because the tension in the studs increases when the pressure in the combustion chamber increases. For the purpose of determination of the ending point in time it is not necessary to convert strain into pressure because the relevant event is the moment of initiation of the pressure rise, and for this it is sufficient to register the moment in which the strain increases.

The combustion sensor can alternatively be a temperature sen- sor mounted in the cylinder cover or in the cylinder lining or wall. The ignition causes a rise in temperature of the gasses contained in the combustion chamber, and this is detectable with a temperature sensor.

Both the pressure increase and the temperature increase are of rather slow acting nature. It is consequently possible to deduct a correc- tion value from the detected moment of temperature or pressure rise in order to obtain the ending point in time. The correction value may be empirically based and stored in the electronic control unit 21 as a predetermined correction value, or it may be manually entered into the system based on observations of how the engine is running. The combustion sensor can alternatively be a light sensor capable of detecting light emitted in the combustion chamber from the ignition of the fuel (preferably L)V light and/or light visible to the human eye). Such a combustion sensor 23 is very fast acting and immediately detects the moment at which ignition begins, and thus detects the above-mentioned ending point in time for the ignition delay with high precision. The combustion sensor 23 for sensing light is in one embodiment mounted as depicted in Fig. 4 where a bore 24 is present in the cylinder cover and a holder 25 having an external threading is screwed into a mating internal threading in a bore leading to the combustion chamber. Holder 25 has a central, through-going bore in which two optical fibres are mounted so that their internal ends are positioned in the combustion chamber, or so closely at the combustion chamber in bore 24 that they have a free view to the contents in the combustion chamber. Of the optical fibres, the one optical fibre 26 is a photo detector fibre, which actually constitutes the combustion sensor. Optical fibre 26 is connected at its outer end to a light sensor device 29, which via a wire 30 transmits a signal to the electronic control unit 21 when the received light intensity is above a threshold value. Light sensor device 29 can be a photo-sensitive diode. The other optical fibre 27 is installed for testing purposes and is connected at its outer end to a light source 28, such as a LED (Light Emitting Diode).

As illustrated in Fig. 3 the cylinder is provided with two combus- tion sensors 23, each comprising said two optical fibres 26, 27 with associated components. The duplicate sensor 23 serves as a back-up sensor in case of failure of the other sensor, and in addition the duplicate sensors also allow the detection system to be a self-testing system that verifies correct functioning of the system. In normal operation the sys- tern is in the measuring mode where each sensor operates individually and delivers signals of detected ignition to the electronic control unit 21. The electronic control unit 21 can compare the two signals, which should be almost identical, and if the signals deviate more than a predetermined threshold value from one another, the electronic control unit 21 may ini- tiate a self-control procedure.

Upon activation of the self-control procedure the functioning of the combustion sensors are checked during the part of the combustion cycle where there is no combustion. The check may e.g. be performed during the compression stroke of the piston. A preferred check proce- dure is the following: A) Light source 28 in one of the sensors is activated to emit light that travels through optical fibre 27 and enters the combustion chamber where the other of the sensors detect the light, thus indicating that both light source and the other sensor are correctly operational. B) Light source 28 in the other of the sensors is activated to emit light that travels through optical fibre 27 and enters the combustion chamber where the one of the sensors detect the light, thus indicating that both light source and the one sensor are correctly operational. C) The detected signals from A) and B) are compared in the electronic control unit 21, and if the deviation between the signals is below a threshold value then correct functioning of both sensors is accepted and normal operation continues. If the deviation is above the threshold value then the detected values are compared to standard values and one of the sensors is accepted as in normal function and the other deactivated and a malfunction signal is issued to an operator. In a preferred embodiment optical fibre 26 is shaped in the end facing the combustion chamber so that the end has at least one planar end surface 31 that forms an angle α with the longitudinal axis of the optical fibre (Fig. 5). The angle α influences the directions β from which the optical fibre can detect light. Line 32 in Fig. 5 illustrates incoming light and the illustration shows how the light is admitted into the optical fibre and reflected at the transition between fibre material and coating 33 on the fibre. The influence of angle α is illustrated in more detail in Fig. 6. In the left hand side of Fig. 6 the end surface is at an angle α = 60° and the optical fibre can detect light in angular sector 34 illustrated by the dark band. At the middle of Fig. 6 the illustration shows that an optical fibre shaped with end surfaces having the angle α = 45° is able to detect light in sectors 35, 35' and 35" covering more than 180°, whereas the illustration in the right hand side of Fig. 6 illustrates that when the end sur- face is at an angle α = 35° then the optical fibre can only detect light in the smaller angular sector 36. The angle α = 45° is preferred.

The ignition delay is determined as the duration between the start of injection and the start of ignition. According to the present invention the ignition delay is utilized to feed the injectors 14 with fuel of low quality in order to produce the desired engine shaft power at low costs. The adjustment of the fuel quality is performed by mixing at least two different fuels into a fuel mixture that on one hand fulfils the requirements of the engine but on the other hand has favourably low costs. One example of a fuel system designed for use with heavy fuel oils is illustrated in Fig. 2. The use of heavy fuel oils places the fuel oil system under certain requirement, because heavy fuel oil is typically not fluid at room temperature. In order to be able to pump the heavy fuel oil it needs to be heated to a temperature elevated well above room tem- perature, such as a temperature of about 70 to 80⁰C. It is furthermore required that the heated heavy fuel oil can circulated through the fuel oil system in amounts in excess of the consumption so that the circulation of heated oil maintains all parts in the system at a sufficiently elevated temperature to avoid blocking due to lacking flowability. And the system has to be designed with a possibility to purge all delicate parts of heavy fuel oil, which is typically performed by switching the operation from heavy fuel oil to marine diesel oil (ISO 8217) or similar oil before the engine is stopped. The fuel oil system has a diesel oil tank 37, which is kept at stand-by during continuous engine operation by setting a valve

38 in a position that blocks tank 37 from the fuel oil system. The tank can be filed with oil through a pipe 39 and be aerated through a vent 40.

A first source of a first fuel is provided in form of a tank 41 which can hold a first heavy fuel oil filled into the tank through a pipe 42. A second source of a fuel conditioner is provided in form of a tank 43 which can hold a second heavy fuel oil filled into the tank through a pipe 44. Vents 40 aerate the tanks. During normal operation of the engine a valve 47 is kept in a position where an outlet pipe 49 from tank 41 is in flow connection with a first set of fuel pumps 48, as indicated by arrow B, and the access to tank 37 is blocked. During normal operation of the engine, valve 38 is kept in a position where an outlet pipe 45 from tank 43 is in flow connection with a second set of fuel pumps 46, as indicated by arrow A, and the access to tank 37 is blocked.

The first fuel is pressurized by the first set of fuel pumps 48 and can flow via a filter 51 to a fuel mixing unit 52 via a conduit 53 provided with a check valve allowing flow only in direction of mixing unit 52. The fuel conditioner is pressurized by the second set of fuel pumps 46 and can flow via a filter 54 to the fuel mixing unit 52 via a conduit 55.

The fuel mixing unit 52 mixes the fuels provided from tanks 41 and 43 into a fuel mixture, which via a conduit 56 is delivered to a third set of fuel pumps 57 that pressurizes the fuel mixture to the pressure required at admission to the fuel dosing devices 10. The fuel system preferably has a fuel heating unit 58 and an additional fuel filter 59 downstream of the third set of fuel pumps. The fuel return line 17 is connected with a pressure regulating valve 60 in the fuel feeding pipe 11 and with a de-aerating device 61 that via a conduit 62 is connected to conduit 56 in order to provide recirculation for the fuel mixture in the fuel feeding pipe 11 and the fuel return line 17. The recirculation ensures that a change in the composi- tion of the fuel mixture is quickly circulated to the fuel oil consuming points where conduits 12 are supplied with the fuel mixture.

The fuel mixing unit 52 is controlled from the electronic control unit 21 which adjusts the mixing ratio of the fuels provided from tanks 41 and 43. The electronic control unit is continuously updated on the current ignition delay via the above-described determination of the duration between the start of injection and the start of ignition. This ignition delay can be obtained from sensors mounted on only a single cylinder in the engine or from sensors mounted on several or all cylinders in the engine.

When the operating conditions or the fuel composition changes so that the ignition delay becomes shorter then the electronic control unit compensates for this change by changing the mixing ratio of the fuel mixture so that the ratio of the less ignitable (and thus poorer and cheaper) fuel is increased. The effect of such an increase is a decrease of the ignitability of the fuel mixture and thus an increase of the ignition delay. For a particular engine the predetermined minimum ignition delay can be set in the electronic control unit. The predetermined ignition delay depends on the engine speed. Generally, when the engine speed is lower then a longer ignition delay is acceptable. For a two-stroke engine having an engine speed of about 80 rpm at 100% engine load (in the present context engine speed is referred to as the speed at 100% engine load) the predetermined value for ignition delay is preferably set at about 3.5 to 3.9 ms. The electronic control unit may in addition to the predetermined value for ignition delay also be provided with a maximum limit for ignition delay. This limit may be set so that proper ignition is assured, possibly with a safety margin based on experience with running with an engine of the actual type. For a two-stroke engine having an engine speed of about 70 rpm the maximum limit for ignition delay is preferably set at about 4.0 ms.

When normal engine operation is to be terminated and the engine stopped, the fuel system can be changed over to operation on die- sel oil from tank 37 by setting valves 38 and 47 in a position opening ac- cess from tank 37 to fuel pumps 46 and 48 and closing valves in outlet pipes 45 and 49.

Other fuels than heavy fuel oil are naturally possible. The fuel conditioner may e.g. be a bio-generated fuel, such as ethanol; and the first fuel may be a very low quality fuel such as fuels containing tar or asphalt. A wide range of other fuels are also possible, in particular because the fuel mixture is adjusted, as the engine is running, to the mixture that is just able to provide the desired ignition properties without using fuel of too high a quality. It is of course possible to have more than two kinds of fuel, such as three, four, five, or more kinds of fuel, which are kept in individual fuel tanks and supplied to the fuel mixture unit via separate pumps and conduits.

The ignitability of the individual fuel can be indicated by the CaI- culated Carbon Aromaticity Index (CCAI) value. This value depends on the viscosity and density of a fuel oil. The CCAI value is calculated by using the following formula:

CCAI = D-81-141 LoglOLoglO (Vk + 0.85) - 483 LoglO ((T + 273)/323) where

Vk = Kinematic Viscosity (mm²/s) at temperature T ⁰C:

D = Density kg/m³ at 15 ⁰C.

It is preferred that the first fuel has a CCIA value higher than the fuel conditioner, so that the fuel is mixed in such a manner that a maximum of fuel with a high CCIA value is spent.

In an embodiment the measurement of the ignition delay is performed at only a single, individual cylinder. This embodiment is easy to implement because only the single cylinder needs to be provided with the combustion sensor and possibly a sensor for detecting start of fuel injection. In another, preferred embodiment several cylinders, such as all cylinders, on the engine are equipped to perform measurement of the ignition delay. This increases reliability of the engine, because a fault condition in one cylinder can be detected when more cylinders can perform measurement of the ignition delay. Modification may be made within the scope of the claims, and it is in particular possible to combine details from the disclosed embodiments into new embodiments.

Claims

P A T E N T C L A I M S

1. A compression ignition internal combustion engine, such as a two-stroke crosshead Diesel engine or a four-stroke Diesel engine, comprising cylinders provided with fuel injectors for injecting fuel directly into combustion chambers in the cylinders, combustion sensors for detecting combustion of fuel in the cylinders, and at least one electronic control unit receiving signals from the combustion sensors, characterized in that a fuel mixing unit is connected with at least a first source of a first fuel and a second source of a fuel conditioner, that based on a detected start of ignition of fuel in the individual cylinder compared with start of fuel injection in said cylinder, the at least one electronic control unit controls the mixing ratio of the first fuel and the fuel conditioner in the fuel mixture delivered by the mixing unit to the fuel injectors, so that the ignitability of the fuel mixture is decreased when the ignition delay is at a predetermined value for ignition delay.

2. An internal combustion diesel engine according to claim 1, characterized in that the first fuel is a fuel of lower costs than the fuel conditioner, which is a fuel of better quality than said first fuel.

3. An internal combustion diesel engine according to claim 1 or 2, characterized in that the first fuel has a CCAI-value that is higher than the CCAI-value of the fuel conditioner.

4. An internal combustion diesel engine according to claim 3, characterized in that the first fuel and the fuel conditioner are heavy fuel oil products.

5. An internal combustion diesel engine according to claim 2, characterized in that the consumption of fuel conditioner is minimized by setting the predetermined value for ignition delay near a predetermined maximum limit for ignition delay.

6. An internal combustion diesel engine according to any of claims 1 to 3, characterized in that at 100% engine load the mixing ratio is so that only the first fuel is supplied to the fuel injectors.

7. An internal combustion diesel engine according to any of claims 1 to 6, characterized in that at 100% engine load the engine speed in the range from 45 rpm to 175 rpm.

8. An internal combustion diesel engine according to claim 1, characterized in that the first fuel is a heavy fuel oil and the fuel conditioner is water.

9. A method of blending a fuel mixture that is injected by fuel injectors directly into combustion chambers in cylinders of a compression ignition internal combustion engine, such as a two-stroke crosshead Diesel engine or a four-stroke Diesel engine, characterized in that based on a detected start of ignition of fuel in the individual cylinder compared with start of fuel injection in said cylinder, the mixing ratio of at least a first fuel and a fuel conditioner in the fuel mixture delivered to the fuel injectors is adjusted to maintain the ignition delay of the fuel mixture longer than a predetermined value for ignition delay.

10. A method of blending a fuel mixture according to claim 9, characterized in that the mixing ratio is adjusted so to varying engine operating conditions that the cost of the fuel mixture is minimized.